Rethinking the Total Cost of Care in AOM and ABS: The Impact of Improved Diagnostic Accuracy and Antibiotic Treatment Where High Efficacy and Adherence Are Achievable

Supplements and Featured Publications, Balancing Outcomes in AOM and ABS: Impact of Efficacy and Adherence and the Implications for Managed, Volume 12, Issue 10 Suppl

Acute otitis media (AOM) and acute bacterial sinusitis (ABS) each pose a considerable financial burden to all involved. Total annual costs for AOM are estimated between $1.4 billion and $4.1 billion,1-5 and in 1996, the total direct healthcare costs for the treatment of sinusitis (defined as primary diagnosis of sinusitis or related airway disease with sinusitis as a comorbid condition) were estimated at $5.8 billion, of which $1.8 billion was attributed to treatment expenditures for children aged 12 years or younger.6 Cost of care is influenced not only by medication costs, but additional direct costs such as office-based provider visits and indirect costs such as caregiver/parent missed work days, which affect employers who provide health insurance to their employees. In cases of treatment failure, the cost of care can far exceed the costs associated with successful therapy.7 Therefore, improved diagnosis and the use of the most effective agents with the highest tolerability profile will improve outcomes and lower the overall cost of therapy.

Treatment success in AOM and ABS is largely dependent on proper diagnosis, the efficacy of agents used to treat these common conditions, and adherence to therapy. Diagnosis of AOM and ABS can be challenging, with many AOM symptoms such as irritability, otalgia, and fever in children also observed in uncomplicated viral upper respiratory tract infections.8,9 Similarly, ABS symptoms mirror symptoms of viral sinusitis. Consequently, AOM and ABS are often overdiagnosed. It is estimated that AOM in children may be overdiagnosed in as many as 50% to 60% of cases.10,11 Often-cited high spontaneous cure rates in AOM (as high as 80%) and ABS (as high as 50%) frequently reflect a high proportion of patients inaccurately diagnosed with these conditions. If the patient never had the disease, spontaneous resolution will be observed.

In industry-sponsored and investigator-initiated research studies, treatment success is determined by clinical and/or microbiologic criteria–specifically, clinical response rates and eradiation rates. In these treatment circumstances, patients are highly motivated, often compensated, and usually carefully monitored to ensure treatment adherence. In contrast, in the less controlled environment of daily clinical practice, adherence to therapy greatly influences treatment outcomes. Therefore, in the clinical practice setting, selection of antibiotic therapy should focus on agents that have the potential to eradicate the causative pathogens and have a tolerability profile such that full course completion of therapy is likely. This need has recently become recognized and incorporated in national guideline recommendations for antibiotic selection options.12,13

Diagnosing AOM and ABS

Diagnostic accuracy in AOM and ABS is essential for optimal medical management. However, these conditions are frequently misdiagnosed. Recent studies of diagnostic accuracy in AOM demonstrate a high rate of diagnostic inaccuracy among pediatricians, family physicians, and nurse practitioners.14,15 In one study, 514 pediatricians viewed 9 different videotaped pneumatic otoscopic examinations of tympanic membranes during a continuing medical education (CME) course and were asked to differentiate between AOM, otitis media with effusion (OME), and a normal tympanic membrane.16 Pediatricians made the correct diagnosis approximately 50% of the time.

Considerable attention was given to improving diagnostic accuracy, especially in AOM. Video-otoscopy was also used to improve diagnostic accuracy of AOM.16 In Rochester, NY, local health maintenance organizations (HMOs) jointly sponsored a CME course that included video-based instruction on improving visual diagnoses for accuracy and differentiation of AOM, OME, and variations of normal.17 Among the studied providers who attended, there was a 19% drop in diagnosis of AOM with a comparative 29% drop in antibiotic use. The providers improved in OME recognition as evidenced by a 17% increase in use of this diagnosis code. The HMOs recovered the cost of attendance for the providers in less than 9 months by the reduction in antibiotic prescriptions. The training's effects were sustained for the entire 16 months of follow-up from August 1, 1997, to January 1, 1999, an indication of the providers' altered behavior.

Explanations of national guideline antibiotic selection recommendations were beneficial because they were a second component of the CME course. However, adoption of guideline recommendations was notoriously poor.18 Included in the instruction was an explanation by an authoritative expert on the reasons and evidence supporting the guideline rationale, interpretation of the applicability to specific patient populations, and making the recommendations locally relevant. A significant shift toward use of guideline-endorsed antibiotics and away from courses of nonguideline-endorsed drugs was observed. The total cost of antibiotic prescriptions was unaffected, but the use of appropriate drugs improved quality and reduced the costs of failed therapy (discussed later).

Tympanocentesis training was the third component included in the CME course. Infant mannequins were used for training this skill. Tympanocentesis can improve diagnostic accuracy because it confirms or refutes the clinical impression made with visual examination. Although some providers were reluctant to perform tympanocentesis in the office setting after the training (possibly out of concern for potential complications or liability), a proportion of the providers went on to perform tympanocentesis in their offices in accordance with national guideline recommendations.

In ABS, the overlap in symptoms with acute viral sinusitis and allergic rhinitis often results in misdiagnosis,11,19 with clinical studies showing that as many as 60% of patients with colds are prescribed antibiotics.10,11 Patients and providers too often associate yellow/green-colored nasal mucus and/or pressure covering the sinuses with ABS. To help differentiate viral sinusitis from ABS, national guidelines recommend that a patient should have yellow/green nasal discharge for 10 or more days, with or without fever, before the diagnosis of ABS can be made.12,20 Few physicians have adopted the recommendations, as a survey of pediatricians and family physicians showed that over 50% treated patients with antibiotics and diagnosed ABS when yellow/green nasal discharge occurred for only 1 day.21

Impact on Managed Care of Accurate Diagnosis and Appropriate Antibiotic Selection for AOM and ABS

An accurate diagnosis of AOM and ABS impacts the cost of care. When a patient is diagnosed with AOM or ABS, antibiotic therapy is usually prescribed. If the patient does not actually have either condition, taking the antibiotic therapy places them at unnecessary risk for adverse effects from treatment. If no improvement is noted, the patient will be considered a treatment failure, and another antibiotic will likely be prescribed. In AOM, recurrent episodes, especially those that do not appear to respond to therapy, lead to referral to an ear, nose, and throat (ENT) specialist and/or possible surgery. In ABS, recurrent infection may result in referral to an ENT specialist for consideration for functional endoscopic sinus surgery. Surgical intervention increases cost of care.

A retrospective study published in 1999 that included more than 22 000 cases of AOM in the pediatric population of a large health plan examined the cost of treatment failure including prescription costs and per-case costs that included clinic visits for each AOM episode.22 For each episode of AOM, amoxicillin was noted as the first choice in therapy; however, selection of amoxicillin declined progressively, because of the greater probability with each additional visit to the clinic that treatment failure occurred. The per-case costs increased from the first to the fifth episodes, from $94 to $134 (see Figure). The increase was attributed to extra office visits, phone consultations, other professional or laboratory services, and surgical intervention, which increased 4-fold in children with 1 episode versus 4 episodes. Overall costs may have been underestimated because visits to hospital outpatient clinics or emergency room visits were not included. These data highlight the economic burden of treatment failure in AOM.

Improved diagnosis of AOM and ABS and the selection of effective therapy will have a positive effect on reducing treatment costs to managed care. Some costs may increase; for example, drug acquisition costs for agents with superior efficacy compared with generic alternatives. However, these costs may be offset by a reduction in total expenditures as a result of a decline in the number of cases of AOM or ABS and the number of treatment failures due to overdiagnosis. In addition, use of effective, guideline-based therapy should reduce the number of treatment failures. In turn, this will decrease the cost of therapy by reducing the cost of managing treatment failures, including expenditures for follow-up office visits, specialist care, and surgical procedures. Rethinking the cost of care to include components such as the cost of unnecessary treatment and treatment failure emphasizes the importance of an accurate diagnosis and selecting effective therapy.

The impact of AOM diagnostic training with an authoritative explanation of the rationale for antibiotic preferences in national guidelines and use of the tympanocentesis procedure in primary care providers' office settings was recently assessed (Casey and Pichichero, 2006 manuscript in progress). The main outcome measured was total cost of care including physician office visits, pharmacy costs, ENT referrals, and surgical costs. Providers were grouped as follows: (1) CME course attendees who adopted strict diagnostic criteria for AOM, use of national guideline-recommended antibiotics for =80% of treatment courses, and tympanocentesis according to guidelines; (2) CME course attendees similar to group 1 but nonadopters of tympanocentesis; (3) non-CME course attendees. Preliminary analysis of the results shows that group 1 had 36% lower total costs of care, primarily derived from fewer office visits and second courses of antibiotics for failed therapy and lower ENT referrals and surgery rates. Group 2 was intermediate with an 18% lower total cost compared to the non-CME course attendees group. It should be noted that although numerous studies show that interactive techniques are the most effective at changing physician care and patient outcomes, traditional methods of CME such as didactic lectures and print pieces were shown to have little or no beneficial effect in changing physician practice.23 Therefore, it is important for educators to pursue programs such as those previously described in their efforts to improve diagnostic accuracy in AOM.

Guideline Recommendations

In AOM, amoxicillin at doses of 80 to 90 mg/kg per day was first recommended in 1999 as first-line therapy by the Centers for Disease Control and Prevention (CDC).24 In cases of amoxicillin treatment failure after 48 to 72 hours, the CDC guideline-recommended high-dose amoxicillin/clavulanate, cefuroxime axetil, or intramuscular (IM) ceftriaxone. In 2004, the American Academy of Pediatrics (AAP) and American Academy of Family Physicians coordinated their endorsement of the CDC guidelines but expanded the choices to include 4 cephalosporins: cefdinir, cefpodoxime, cefuroxime, and ceftriaxone.

For ABS, in the AAP and Sinus and Allergy Health Partnership Panel national guidelines,12,20 amoxicillin was recommended at doses of 45 to 90 mg/kg per day.12 For amoxicillin failures, those who were treated with antimicrobial therapy, had moderate-to-severe illness, or attended day care, high-dose amoxicillin/clavulanate was recommended.12 Cephalosporin alternatives endorsed were cefdinir, cefuroxime, cefpodoxime, and ceftriaxone.12 These recommendations are summarized in the Table.

New Data Cause Reconsideration of Antibiotic Recommendations

Haemophilus influenzae

Streptococcus pneumoniae



S pneumoniae

To minimize the risk of treatment failure, all guidelines suggest that empiric therapy for AOM and ABS should be effective against the most common causative pathogens, specifically , , and . Until recently, was the major causative pathogen in AOM, with penicillin nonsusceptible pneumococcal (PNSP) strains an increasing concern in the 1990s.25-27 However, with the release of the pneumococcal conjugate vaccine (PCV7) in 2000, a shift occurred in the proportion of infections caused by each of these pathogens.



S pneumoniae

H influenzae


H influenzae

H influenzae


H influenzae


The PCV7 vaccine, which targeted the 5 pneumococcal subtypes most commonly observed in AOM that also accounted for most PNSP strains, eliminated a portion of these strains in the pediatric population.28,29 At the same time, an increase occurred in the proportion of infections caused by . Two prospective, community based studies documented this shift, both showing a significant decrease in AOM caused by from 48% prior to the release of PCV7 to 31% in the post-PCV7 era.28,29 In one study, conducted in rural Kentucky, the percentage of infections caused by nontypable increased significantly from 41% to 56% (= .01) with the percentage of ß-lactamase-producing strains among isolates increasing from 56% in the pre-PCV7 era to 64% in the post-PCV7 era.28 The second study, conducted in suburban Rochester, NY, in patients with persistent AOM or AOM treatment failures, also noted a significant increase in isolates (= .012) and b-lactamase-producing isolates (= .04) in the post-PCV7 era.29

Data documenting the effect of PCV7 on the microbiology of AOM in adults and ABS in children and adults are lacking. However, because the bacteria that cause AOM and acute sinusitis infection in children and adults are generally the same as those that cause AOM in children,30 it is reasonable to suggest that the pathogen shift that has occurred in AOM in children has also occurred in AOM in adults and ABS.

S pneumoniae

H influenzae



H influenzae

M catarrhalis

H influenzae

H influenzae

Treatment selection in AOM in the various national guidelines was based primarily on antibacterial activity against , especially PNSP. and were considered secondary and tertiary pathogens of concern, respectively. The shift in the distribution of causative pathogens as a result of PCV7 requires that agents selected for therapy be active against ß-lactamase-producing nontypable and , because these organisms account for almost two thirds of pathogens isolated in AOM in patients who have received 3 or 4 doses of PCV7.28,29 Whereas high-dose amoxicillin is still a cost-effective choice for uncomplicated AOM and ABS, for patients with persistent or recurrent infection, antibiotic selection should focus on a higher likelihood of ß-lactamase-producing nontypable . Because amoxicillin does not adequately eradicate ß-lactamase-producing organisms and would not be effective therapy, a focus on agents with superior clinical cure and eradication rates for nontypable , including the ß-lactamase-producing type, becomes important.

Adherence Considerations

Efficacy alone cannot predict treatment outcomes. Even agents with superior clinical cure and eradication rates will not be effective in clinical practice if adherence to therapy is lacking. Adherence considerations include duration of therapy, taste, dosing, and tolerability to improve adherence, and the likelihood of treatment success. Nonadherence frequently results in failed therapy, persistence of infection, and morbidity.31

An AAP physician survey that asked 800 pediatricians what they believed were the most important reasons why infants and children fail to take medication as prescribed noted that frequency of dosing, unpleasant taste or palatability, side effects, and duration of therapy were each highly associated with treatment adherence (AAP Periodic Survey #44. January-May 2000; 803 US pediatrician responders). Not surprisingly, factors such as taste, duration of therapy, and dosing schedule influence physicians' and parents' preferences for therapy.32,33 In a survey of 400 parents and 100 pediatricians that asked parents to comment on their knowledge and concerns about antibiotic therapy,34 41% of parents ranked the effectiveness of the antibiotic as their primary concern, 37% chose taste as their primary concern, and 33% chose dosing schedule. Pediatricians surveyed responded that they were most likely to think that parents were concerned about dosing schedule (63%), followed by effectiveness (53%). A total of 32% thought that parents were concerned about taste.

In a brief review of AOM and its treatment options, tolerability and ease of antibiotic administration were viewed as primary considerations in the antibiotic selection process.35 It was further noted that adherence to prescribed therapy may be improved with increased attention to these factors. Fewer doses per day also has a positive effect on adherence.36 This effect was also seen in other disease states.37

Shorter duration of therapy has an extremely positive impact on adherence because it recognizes the reality of patient behavior. Most patients take antibiotics until they feel better (usually 4-6 days), then stop. A study asked 86 physicians and healthcare personnel to evaluate antibiotics on taste and other factors likely to influence overall satisfaction with therapy.38 Shorter duration of therapy (ie, 5-day vs standard 10-day therapy) was viewed as preferable to fewer doses per day, except when the shorter duration of therapy requires agents to be given 3 or 4 times a day compared with once daily.38 Shorter duration of therapy is therefore important to consider for promoting patient adherence, especially because 5 days of therapy was proven as effective as 10 days of treatment in AOM in a review of 27 clinical trials involving almost 7000 patients.39 Additional clinical studies have confirmed the equivalent efficacy of shorter courses of therapy compared with standard 10-day regimens.35,40-42

Taste or palatability plays an important role in treatment satisfaction and adherence to therapy for pediatric patients. Among pediatric patients, the taste of the medication may be a motivating factor for adherence to therapy.43,44 In the same study that asked physicians and healthcare personnel to rank factors likely to affect parent and patient satisfaction, taste was given the highest ranking among appearance, smell, texture, taste, and aftertaste.38 Consequently, a better tasting product that improves adherence may justify a more expensive acquisition cost and still be considered cost-effective.44


Rethinking the cost of care for AOM and ABS requires a consideration of treatment costs that include not just the cost of antimicrobial therapy but also the cost of unnecessary treatment and treatment failures. An accurate diagnosis of AOM and ABS is key to reducing overdiagnosis and promoting the proper selection of antibiotics. Although the most appropriate therapy may result in higher drug acquisition costs, the overall cost of therapy will likely decline because of fewer treatment failures and subsequently less need to refer patients to specialists for consultation or surgical procedures.

To promote positive clinical outcomes, agents selected should be effective and promote adherence. Clinical trials can establish the efficacy of antibiotic therapy based on clinical cure rates and eradication rates; however, even the most effective agent can result in treatment failure if the patient refuses to adhere to therapy. Lack of adherence to therapy places the patient at risk for treatment failure and long-term consequences, including recurrent infection or, in the case of AOM, hearing loss. Adherence-promoting agents have favorable taste, appearance of the suspension, duration of therapy (5 vs 10 days), and dosing intervals. In addition, patient and parent or caregiver preference based on these factors must also be considered when selecting therapy. Finally, side effects (ie, diarrhea and gastrointestinal pain) may result in treatment discontinuation and thus treatment failure. Therefore, the ideal agent for the treatment of AOM and ABS should be both clinically effective and possess a favorable adherence.

We would like to thank the University of the Sciences in Philadelphia for

editorial assistance in the development of these articles.

Address Correspondence to: Michael E. Pichichero, MD, University of Rochester Medical Center, Elmwood Pediatric Group, 601 Elmwood Avenue, Box 672, Rochester, NY 14642. E-mail: